Brit. 3. Psychiat. (1966), 112, 819-825

Psychiatric Disorders in Foster Home Reared Children of

Schizophrenic Mothers*

By LEONARD L. HESTON

INTRODUCTION

The place of genetic factors in the aetiologyof schizophrenia remains disputed. Severalsurveys have demonstrated a significantly higherincidence of the disorder in relatives of schizophrenic persons as compared to the generalpopulation. Furthermore, the closer the relationship, the higher the incidence of schizophrenia.The studies of Kallmann ( i 938) and Slater(I 953) are especially significant and the research

in this area has been thoroughly reviewed byAlanen (1958).

Although the evidence for a primarily geneticaetiology of schizophrenia is impressive, analternative explanation—that schizophrenia isproduced by a distorted family environmenthas not been excluded. A close relative who isschizophrenic can be presumed to produce adistorted interpersonal environment and thecloser the relationship the greater the distortion.

This study tests the genetic contribution toschizophrenia by separating the effects of anenvironment made “¿�schizophrenogenic― by the

ambivalence and thinking disorder of a schizophrenic parent from the effects of genes fromsuch a parent. This is done by comparing agroup of adults born to schizophrenic motherswhere mother and child were permanentlyseparated after the first two postpartum weekswith a group of control subjects.

SELECTION OF SUBJECTS

The Experimental subjects were born betweenI 9 I 5 and i@ to schizophrenic mothers con

fined to an Oregon State psychiatric hospital.Most of the subjects were born in the psychiatrichospital ; however, hospital authorities encour

‘¿�pThis research was supported by the Medical Research

Foundation of Oregon.

aged confinement in a neighbouring generalhospital whenever possible, in which case thechildren were delivered during brief furloughs.All apparently normal children born of suchmothers during the above time span wereincluded in the study if the mother's hospitalrecord ( I ) specified a diagnosis of schizophrenia,dementia praecox, or psychosis ; (2) containedsufficient descriptions of a thinking disorder orbizarre regressed behaviour to substantiate thediagnosis ; (3) recorded a negative serologic testfor syphilis and contained no evidence ofcoincident disease with known psychiatricmanifestations ; and (@) contained presumptiveevidence that mother and child had beenseparated from birth. Such evidence typicallyconsisted of a statement that the mother hadyielded the child for adoption, a note that thefather was divorcing the mother, the continuedhospitalization of the mother for several years,or the death of the mother. In practice theserequirements meant that the mothers as a

group were biased in the direction of severe,chronic disease. No attempt was made toassess the psychiatric status of the father;however, none were known to be hospitalpatients. The 74 children ascertained as abovewere retained in the study if subsequent recordsearches or interviews confirmed that the childhad had no contact with its natural mother andnever lived with maternal relatives. (The latterrestriction was intended to preclude significantexposure to the environment which might haveproduced the mother's schizophrenia.)

All of the children were discharged from theState hospital within three days of birth (inaccordance with a strictly applied hospitalpolicy) to the care of family members or tofoundling homes. The records of the child careinstitutions made it possible to follow many

8ig

—¿�ExperimentalControlMaleFemaleMaleFemaleNumber

. . . .Died, infancy or childhoodLost to follow-up . ..

.

. .

. ..

.

. .

. ..

.

. .

. .33

325 623325

53@ifl!:i

Groups . . . .. .. .. .30I 733I 7

820 PSYCHIATRIC DISORDERS IN FOSTER HOME REARED CHILDREN

subjects through their early life, including, forsome, adoption. The early life of those subjectsdischarged to relatives was less completelyknown, although considerable information wasdeveloped by methods to be described.

Sixteen subjects were dropped because ofinformation found in foundling home records;

6 children, 4 males and 2 females, died in earlyinfancy. Ten others were discarded, 8 becauseofcontact with their natural mother or maternalrelatives, one because of multiple gastrointestinal anomalies, and one because nocontrol subject whose history matched thebizarre series of events that complicated theExperimental subject's early life could be found.The remaining 58 subjects comprise the finalExperimental group.

A like number of control subjects, apparentlynormal at birth, were selected from the recordsof the same foundling homes that received someof the Experimental subjects. The controlsubjects were matched for sex, type of eventualplacement (adoptive, foster family, or institutional), and for length of time in child careinstitutions to within ±‘¿�o per cent. up to 5 years.

(Oregon State law prohibited keeping a childin an institution more than five years. Subjectsin institutions up to this maximum were countedas “¿�institutionalized―regardless of final placement.) Control subjects for the Experimentalchildren who went to foundling homes wereselected as follows : When the record of anExperimental subject was located, the admissionnext preceding in time was checked, then thenext subsequent, then the second preceding andso on, until a child admitted to the home withina few days of birth and meeting the abovecriteria was found. Those Experimental subjectswho were never in child care institutions were

matched with children who had spent less thanthree months in a foundling home. The abovemethod of selection was used with the recordsearch beginning with an Experimental child'syear of birth. The above restrictions regardingmaternal contacts were applied to the controlgroup. Oregon State psychiatric hospital recordswere searched for the names of the naturalparents (where known) of the control subjects.In two cases a psychiatric hospital record waslocated and the children of these persons werereplaced by others. All of the children went tofamilies in which both parental figures werepresent.

Exact matching was complicated by the subsequent admission of several subjects to otherchild care institutions and by changes of fosteror even adoptive homes. However, these disruptions occurred with equal frequency and

intensity in the two groups and are consideredrandom.

Table I gives the sex distribution of thesubjects and the causes of further losses. Fifteenof the 74 Experimental subjects died beforeachieving school age. This rate is higher thanthat experienced by the general populationfor the ages and years involved, but not significantly so.

FOLLOW-UP METHOD

Starting in i 964, it proved possible to locateor account for all of the original subjects exceptfive persons, all females. During this phase ofthe research, considerable background information of psychiatric import was developed.The records of all subjects known to policeagencies and to the Veterans' Administrationwere examined. Retail credit reports were

T@rn@zI

‘¿�U

821BY LEONARD L. HESTON

obtained for most subjects. School records,civil and criminal court actions, and newspaperfiles were reviewed. The records of all publicpsychiatric hospitals in the three West CoastStates were screened for the names of thesubjects and the records located were reviewed.Enquiries were directed to psychiatric facilitiesserving other areas where subjects were living,

and to probation departments, private physicians, and various social service agencies withwhich the subjects were involved. Finally,relatives, friends and employers of most subjectswere contacted.

In addition to information obtained from theabove sources, for most subjects the psychiatricassessment included a personal interview, aMinnesota Multiphasic Personality Inventory(MMPI), an I.Q. test score, the social class ofthe subject's first home, and the subject'scurrent social class. As the subjects were located,they were contacted by letter and asked toparticipate in a personal interview. The interview was standardized, although all promisingleads were followed, and was structured as ageneral medical and environmental questionnaire which explored all important psychosocial dimensions in considerable depth. Nearlyall of the interviews were conducted in thehomes of the subjects, which added to the rangeof possible observations. The short form of theMMPI was given after the interview. The

. results ofan I.Q. test were available from school

or other records for nearly all subjects. If a testscore was not available, the Information,Similarities, and Vocabulary subtests of theWechsler Adult Intelligence Scale (WAIS) wasadministered and the I.Q. derived from theresults. Two social class values were assignedaccording to the occupational classificationsystem of Hollingshead (1958). One value wasbased on the occupation of the father orsurrogate father of the subject's first family atthe time of placement, and a second on thesubject's present occupational status or, formarried females, the occupation of the husband.The social class values move from i to 7 withdecreasing social status.

All of the investigations and interviews wereconducted by the author in 14 States and inCanada.

EVALUATION OF SUBJECTS

The dossier compiled on each subject,excluding genetic and institutional information,was evaluated blindly and independently bytwo psychiatrists. A third evaluation was madeby the author. Two evaluative measures wereused. A numerical score moving from ioo to owith increasing psycho-social disability wasassigned for each subject. The scoring was basedon the landmarks of the Menninger MentalHealth-Sickness Rating Scale (MHSRS)(Luborsky, 1962). Where indicated, the ratersalso assigned a psychiatric diagnosis from theAmerican Psychiatric Association nomenclature.

Evaluations ofg7 persons were done. Seventytwo subjects were interviewed. Ofthe remaining25 persons, six refused the interview (7 6 per

cent. of those asked to participate), eight weredeceased, seven are inaccessible (active inArmed Forces, abroad, etc.), and four were notapproached because of risk of exposure of thesubject's adoption. It did not seem reasonableto drop all of these 25 persons from the study,since considerable information was available formost of them. For instance, one man was killedin prison after intermittently spending mostof his life there. His behavioural and socialrecord was available in prison records plus theresults of recent psychological evaluations.A man who refused the interview was a known,overt, practising homosexual who had a recentfelony conviction for selling narcotics. Allpersons in the Armed Forces were knownthrough letters from their Commanding Officersor medical officers to have been serving honourably without psychiatric or serious behaviouralproblems. One 21-year-old man, the leastknown of any of the subjects, had been inEurope for the preceding 18 months in anuncertain capacity. He is known to havegraduated from high school and to have noadverse behavioural record. In a conferencethe raters agreed that it would be misleadingto discard any cases, and that all subjectsshould be rated by forced choice.

The MHSRS proved highly reliable as ameasure of degree of incapacity. The IntraclassCorrelation Coefficient between the scoresassigned by the respective raters was o@

822

indicating a high degree of accuracy. Asexpected, several differences arose in the assignment of specific diagnoses. In disputed casesa fourth psychiatrist was asked for an opinionand differences were discussed in conference.The only differences not easily resolved involveddistinctions such as obsessive-compulsive neuro

sis versus compulsive personality or mixedneurosis versus emotionally unstable personality.All differences were within three diagnosticcategories : psychoneurotic disorders, personalitytrait or personality pattern disturbances. Theraters decided to merge these categories intoone : “¿�neuroticpersonality disorder―. Thiscategory included all persons with MHSRSscores less than 75—the point on the scale wherepsychiatric symptoms become troublesomewho received various combinations of the abovethree diagnoses. In this way, complete agreement on four diagnoses was achieved: schizophrenia, mental deficiency, sociopathic personality, and neurotic personality disorder.

PSYCHIATRIC DISORDERS IN FOSTER HOME REARED CHILDREN

One mental defective was also diagnosedschizophrenic and another sociopathic. Onlyone diagnosis was made for all other subjects.

RESULTS

Psychiatric disability was heavily concentrated in the Experimental group. Table IIsummarizes the results.

The MHSRS scores assess the cumulativepsycho-social disability in the two groups. Thedifference is highly significant with the Experimental group, the more disabled by thismeasure. However, the difference is attributableto the low scores achieved by about one-half(26/47) ofthe Experimental subjects rather than

a general lowering of all scores.

The diagnosis of schizophrenia was based ongenerally accepted standards. In addition to theunanimous opinion of the three raters, allsubjects were similarly diagnosed in psychiatrichospitals. One female and four males com

TAm.n Ii

ExactControl Experimental Probability

Number . . . . . . . . ..Male . . . . . . . . . . ..Age, Mean . . . . . . . . ..Adopted . . . . . . . . ..MHSRS, Mean (Total group mean = 72 8, S.D.Schizophrenia (Morbid Risk = :6 . 6%) ..Mental Deficiency (T.Q. < 70) . . ..Sociopathic Personality . . . . ..Neurotic Personality Disorder . . ..Persons spending > I year in Penal or Psychiatric

Total years Institutionalized . . ..Felons . . . . . . . . . . ..Armed Forces, Number Serving . . ..Armed Forces, Number Discharges, Psychiatric or

Behavioural . . . . . . . . ..Social Group, First Home, Mean . . . . ..Social Group, Present, Mean . . . . ..I.Q., Mean . . . . . . . . . . ..Years School, Mean . . . . . . . . ..Children, Total . . . . . . . . ..Divorces, Total . . . . . . . . ..Never Married, > 30 Years Age . . . . ..

5°3336.3‘¿�980@I

0

02

72

‘¿�52

‘¿�7

473°35@8226@.2

549

‘¿�3‘¿�I

I 12

72!

8

4.5

5.494@0

7'69

= 18.4)

Institution

0@ ooo600240@O52001700520@ oofi

0@ 054

0021. . 42

. . 4.7

. . 103.7

. . 124

. . 84

.. 7

.. 44'

One mental defective was also schizophrenic.Another was sociopathic.Considerable duplication occurs in the entries below Neurotic Personality Disorder.* Fisher Exact Probability Test.

823BY LEONARD L. HESTON

prised the schizophrenic group. Three werechronic deteriorated patients who had beenhospitalized for several years. The other twohad been hospitalized and were taking antipsychotic drugs. One of the latter persons wasalso mentally deficient : a brief history of thisperson follows.

A farm labourer, now 36 years old, was in an institutionfor mentally retarded children from age 6-i6. Several I.Q.tests averaged 62. He was discharged to a family farm,where he worked for the next i 6 years. Before his hospitalization at age 32 he was described as a peculiar butharmless person who was interested only in his bankaccount : he saved $5,500 out of a salary averaging $gooper year. Following a windstorm that did major damageto the farm where he worked he appeared increasinglyagitated. Two days later he threatened his employer witha knife and accused him of trying to poison him. A courtcommitted him to a psychiatric hospital. When admitted,he talked to imaginary persons and assumed a posture ofprayer for long periods. His reponses to questions wereincoherent or irrelevant. The hospital diagnosis wasschizophrenic reaction. He was treated with phenothiazine drugs, became increasingly rational, and wasdischarged within a month. After discharge he returnedto the same farm, but was less efficient in his work andspent long periods sitting and staring blankly. He hasbeen followed as an out-patient since discharge, has takenphenothiazine drugs continuously, and anti-depressantsoccasionally. This man exhibited almost no facial expression. His responses to questions, though relevant, weregiven after a long and variable latency.

The age-corrected rate for schizophrenia isI 6 . 6 per cent., a finding consistent with

Kallmann's i6 .4 per cent. (Weinberg's shortmethod, age of risk 15—45years). Hoffman( I 92 I ) and Oppler ( i 932) reported rates of from7 to I0@ 8 per cent. of schizophrenia in childrenof schizophrenics. No relationship between theseverity and sub-type of the disease in themother-child pairs was evident.

Mental deficiency was diagnosed when asubject's I.Q. was consistently less than 70.All of these persons were in homes for mentaldefectives at some time during their life andone was continuously institutionalized. HisI.Q. was 35. The other mentally deficientsubjects had I.Q.s between 50 and 65. Nohistory of CNS disease or trauma of possiblecausal importance was obtained for any of thesesubjects. The mothers of the mentally defectivesubjects were not different from the othermothers and none were mentally defective.

Three behavioural traits were found almostexclusively within the Experimental group.These were : ( i) significant musical ability,7 persons; (2) expression of unusually strongreligious feelings, 6 persons ; and (@) problemdrinking, 8 persons.

The results with respect to the effects ofinstitutional care, social group, and type ofplacement will be discussed in a later paper.None of these factors had measurable effectsOn the outcome.

DIscUssIoN

The results of this study support a geneticaetiology of schizophrenia. Schizophrenia wasfound only in the offspring of schizophrenicmothers. The probability of this segregationbeing effected by chance is less than o@ 025.Furthermore, about one-half of the Experimental group exhibited major psycho-socialdisability. The bulk of these persons haddisorders other than schizophrenia which werenearly as malignant in effect as schizophreniaitself. An illustration is provided by the 8 of2 I Experimental males who received psychiatric

or behavioural discharges from the armedservices. If three subjects who were rejected forservice for the same reasons are added, the ratiobecomes I I : 24, or essentially I : 2. Only threeof these I I subjects were schizophrenic and oneschizophrenic served honourably. Kallmann's( 1938) rate for first degree relatives and Slater's( I 953) for dizygotic twins of schizophrenicpersons who developed significant psycho-socialdisability not limited to schizophrenia areslightly lower, though in the same range, asthose found in the present study.

The association of mental deficiency withschizophrenia has been reported by Hallgrenand Sjagren (i 959) who noted an incidence oflow-grade mental deficiency (I.Q. < 50—55)inschizophrenic subjects of about i o@ 5 per cent.Kallmann (1938) found from 5—10per cent.mental defectives among his descendants ofschizophrenic persons, but did not consider thefinding significant. The association of mentaldeficiency with schizophrenia—if such anassociation exists—remains uncertain.

824 PSYCHIATRIC DISORDERS IN FOSTER HOME REARED CHILDREN

Two sub-groups of persons within the impaired one-half of Experimental subjects cxhibited roughly delineable symptom-behaviourcomplexes other than schizophrenia or mentaldeficiency. The personalities of the personscomposing these groups are described inaggregate below.

The first group is composed of subjects whofit the older diagnostic category, “¿�schizoidpsychopath―. This term was used by Kalhnann(1938) to describe a significant sub-group of his

relatives of schizophrenic persons. Eight malesfrom the present study fall into this group,all of whom received a diagnosis of sociopathicpersonality. These persons are distinguished byanti-social behaviour of an impulsive, illogicalnature. Multiple arrests for assault, battery,poorly planned impulsive thefts dot their policerecords. Two were homosexual, four alcoholic,and one person, also homosexual, was anarcotics addict. These subjects tended to livealone—only one was married—in deterioratedhotels and rooming houses in large cities, andlocating them would have been impossiblewithout the co-operation of the police. Theyworked at irregular casual jobs such as dishwasher, race-track tout, parking attendants.When interviewed they did not acknowledge orexhibit evidence of anxiety. Usually secretiveabout their own life and circumstances, theyexpressed very definite though general opinionsregarding social and political ills. In spite oftheir suggestive life histories, no evidence ofschizophrenia was elicited in interviews. Nosimilar personalities were found among thecontrol subjects.

A second sub-group was characterized byemotional lability and may correspond to theneurotic sibs of schizophrenics described byAlanen (1963). Six females and two males fromthe Experimental group as opposed to twocontrol subjects were in this category. Thesepersons complained of anxiety or panic attacks,hyper-irritability, and depression. The mostfrequent complaint was panic when in groupsof people as in church or at parties, which wasso profoundly uncomfortable that the subjectwas forced to remove himself abruptly. Mostsubjects described their problems as occurringepisodically ; a situation that they might tolerate

with ease on one occasion was intolerable onanother. The woman reported life-long difficulty with menses, especially hyper-irritabilityor crying spells, and depressions coincident withpregnancy. These subjects described themselvesas “¿�moody―,stating that they usually could notrelate their mood swings to temporal events.Four such subjects referred to their strongreligious beliefs much more frequently thanother respondents. Psychophysiological gastrointestinal symptoms were prominent in fivesubjects. The most frequent diagnoses advancedby the raters were emotionally unstable personality and cyclothymic personality, withneurosis a strong third.

Of the 9 persons in the control group whowere seriously disabled, 2 were professionalcriminals, careful and methodical in their work,2 were very similar to the emotionally labile

group described above, one was a compulsivephobia-ridden neurotic, and 4 were inadequateor passive-aggressive personalities.

The 2 I Experimental subjects who exhibitedno significant psycho-social impairment werenot only successful adults but in comparisonto the control group were more spontaneouswhen interviewed and had more colourfullife histories. They held the more creative jobs:musician, teacher, home-designer ; and followedthe more imaginative hobbies : oil painting,music, antique aircraft. Within the Experimental group there was much more variabilityof personality and behaviour in all socialdimensions.

SUMMARY

This report compares the psycho-socialadjustment of 47 adults born to schizophrenicmothers with 50 control adults, where allsubjects had been separated from their naturalmothers from the first few days of life. Thecomparison is based on a review of school,police, veterans, and hospital, among severalother records, plus a personal interview andMMPI which were administered to 72 subjects.An I.Q. and social class determination werealso available. Three psychiatrists independently rated the subjects.

1

I

825BY LEONARD L. HESTON

County Probation Department; Robert Tyler, ResearchInformation Director, California Bureau of Corrections;Evan Iverson, State of Washington, Department ofInstitutions; Anthony Hordern, Chief of Research,California Department of Mental Hygiene; CaptainGeorge Kanz, Oregon State Police; J. S. Gleason,Administrator, Veterans' Administration; and Lt.-GeneralLeonard D. Heaton, Rear-Admiral E. C. Kenney, andMajor-General R. L. Bohannon, Chief Medical Officersofthe Army, Navy and Air Force respectively.

Renate Whitaker, University of Oregon MedicalSchool, and Eliot Slater and James Shields of thePsychiatric Genetics Research Unit, Maudsley Hospital,London, reviewed the manuscript and made manyhelpful suggestions.

Finally, I wish gratefully to acknowledge the contribution made by the subjectsof this research project, most ofwhom freely gave ofthemselves in the interest of furtheringmedical science.

REFERENCKS

ALANEN, Y. 0. (1958). “¿�The mothers of schizophrenic

patients.― Acta psychiat. ricurol. scand., Suppi. i 227.—¿�, REKOLA, J., STAVEN, A., TUOVINEN, M.,@

K., and Ru@r@s.nzN,B. (i96@). “¿�Mentaldisorders inthe siblings of schizophrenic patients.―Ada psychiat.scand., Suppl. 169, 39, 167.

HALLGREN, B., and SJöGREN, T. (1959). “¿�Aclinical andgenetico-statistical study of schizophrenia and lowgrade mental deficiency in a large Swedish ruralpopulation.― Ada psychiat. neurol. scand., Suppl. 140.Vol. 35.

HOFFMAN, H. (1921). “¿�Studien über Vererbung undEntstehung geistiger Storungen. II. Die Nachkommenschaft bei endogenen Psychosen.― Berlin:Springer.

HOLLINGSHEAD, A. B., and REDLICH, F. C. (1958).Social Class and Menial illness : A Community Study.New York: J. Wiley.

KALLMANN, F. J. (1938). The Geneti€s of Schizophrenia.New York: J. J. Augustin.

LUBORSKY, L. (1962). “¿�Clinicians' judgements of mentalhealth: a proposed scale.―Arch. gen. Psychiat. (Chic.),7, 407.

OPPLER, W. ( i 932). “¿�Zum Problem der Erbprognose

bestimmung.― @.Neurol., i.@i, 549—616.

SLATER, E., with SHIELDS, J. ( i 953). Psychotic and neuroticillnesses in twins. Medical Research Council SpecialReport Series No. 278. London: H.M. StationeryOffice.

The results were:( I ) Schizophrenic and sociopathic person

ality disorders were found in those persons bornto schizophrenic persons in an excess exceedingchance expectation at the o@ 05 level of probability. Five of47 persons born to schizophrenicmothers were schizophrenic. No cases of schizophrenia were found in 50 control subjects.

( 2) Several other comparisons, such as per

sons given other psychiatric diagnoses, felons,and persons discharged from the Armed Forcesfor psychiatric or behavioural reasons, demonstrated a significant excess of psycho-socialdisability in about one-half of the persons bornto schizophrenic mothers.

(3) The remaining one-half of the personsborn to schizophrenic mothers were notablysuccessful adults. They possessed artistic talentsand demonstrated imaginative adaptations tolife which were uncommon in the control group.

ACKNOWLEDGMENTS

The author is greatly indebted to Drs. Duane D.Denney, Ira B. Pauly, and Arlen Quasi who evaluated thecase histories and provided invaluable advice andencouragement. Drs. Paul Blachly, John Kangas, HaroldOsterud, George Saslow, and Richard Thompson providedadvice and/or facilities which greatly contributed to thesuccess of the project. All of the above are faculty or staffmembers of the University of Oregon Medical School.

This research could not have been completed withoutthe splendid co-operation of numerous officials of variousagencies who provided indispensable information. I amespecially indebted to the following: Dean R. Mathews,Waverly Baby Home; Elda Russell, Albertina KerrNurseries; Stuart R. Stim.mel and Esther Rankin, Boys'and Girls' Aid Society of Oregon; Reverend Morton E.Park, Catholic Charities; George K. Robbins, JewishFamily and Child Services; Miss Marian Martin, State ofOregon, Department of Vital Statistics, all of Portland,Oregon. Drs. Dean K. Brooks,E. I. Silk,Russel M. Guiss,J. M. Pomeroy,Superintendents ofOregon State Hospital,Eastern Oregon State Hospital, Dammasch State Hospital,and Oregon Fairview Home, respectively. David G.Berger, Research Coordinator, Oregon State Board ofControl ; Stewart Adams, Research Director, Los Angeles

Leonard L. Heston, M.D., Resident in Psychiatry, University of Oregon Medical School. Currently GuestWorker, Psychiatric Genetics Research Unit, Maudsley Hospital, London

(Received II Xovember, 1965)

10.1192/bjp.112.489.819Access the most recent version at DOI: 1966, 112:819-825.BJP 

LEONARD L. HESTONSchizophrenic MothersPsychiatric Disorders in Foster Home Reared Children of

Referenceshttp://bjp.rcpsych.org/content/112/489/819#BIBLThis article cites 0 articles, 0 of which you can access for free at:

permissionsReprints/

permissions@rcpsych.ac.ukto To obtain reprints or permission to reproduce material from this paper, please write

to this article atYou can respond /letters/submit/bjprcpsych;112/489/819

from Downloaded

The Royal College of PsychiatristsPublished by on October 5, 2017http://bjp.rcpsych.org/

http://bjp.rcpsych.org/site/subscriptions/ go to: The British Journal of PsychiatryTo subscribe to